Computer Vision
MicroSDF: Microfacet-Driven Hybrid Neural SDFs for Mixed-Reflectance Surface Reconstruction
Lejia Ye, Yuhua Xu, Yulan Guo et al. (+2 more)
Accurate 3D reconstruction in real-world environments remains a significant challenge due to the coexistence of reflective and non-reflective surfaces, which pose distinct modeling demands. Existing methods often treat these surface types separately, limiting their generalizability and physical plausibility. To bridge this gap, we propose MicroSDF, a novel neural implicit framework that facilitates geometry and reflectance modeling through microfacet theory. Our approach incorporates three core innovations: (i) a microfacet-guided geometry model that extracts multi-scale surface normals (macroscopic and microfacet) from a signed distance field (SDF), regularized by a proposed microfacet normal consistency loss to enforce physically plausible surface orientations; (ii) an enhanced dual-branch color model, where the specular branch leverages the microfacet normals to model high-frequency reflectance, and the vanilla branch, unlike prior works, uses reflection direction (instead of viewing direction) to better model diffuse and low-frequency specular components; and (iii) a detection-guided color blending strategy that adaptively fuses the color outputs based on reflection priors, providing more physically intuitive blending than implicitly learned blending weights. Combined with a tailored multi-stage optimization scheme, the proposed MicroSDF achieves robust and high-fidelity reconstruction across reflective and non-reflective surfaces. Extensive experiments on DTU, Shiny Blender, Ref-NeRF, and DeepVoxels
datasets demonstrate state-of-the-art performance, establishing a new direction for physically grounded neural reconstruction.
Read the full abstract ▾
All authors Lejia Ye, Yuhua Xu, Yulan Guo, Lian Xu, Xiaohu Zhang
Nuclear Experiment
Probing the production mechanisms of virtual photons at RHIC
STAR Collaboration
We report the measurement of direct virtual
photon ($\gamma^*_\text{dir} \rightarrow e^+e^-$) in four Au+Au collision energies at $\sqrt{s_\text{NN}}$ = 14.6–54.4 GeV. The $\gamma^*_\text{dir}$ yields scale with charged-particle multiplicity ($(dN_{\mathrm{ch}}/d\eta)^{\alpha}$) as $\alpha=1.43 \pm 0.04 \text{(stat.)} \pm 0.03\text{(sys.)}$, independent of collision energies and centralities, exceeding hadronic expectations and indicating notable contribution from QGP phase. At $\sqrt{s_{\rm NN}}=$ 54.4 GeV, the direct virtual
photon mass distributions are further studied over $0.1 < M_{ee} < 1.1$ GeV/$c^2$, and are found to evolve from exponential to power-law form with increasing $p_\text{T}$, signaling a shift from "blackbody" radiation to emission governed by Kroll-Wada formalism.
Read the full abstract ▾
Artificial Intelligence
MCAIR: Monitoring and Correcting Agent Unfaithfulness via Internal Representations
Yan Xu, Dandan Song, Luan Zhang et al. (+8 more)
Large
Language Model (LLM)-based
agents have demonstrated remarkable capabilities in completing complex tasks through multi-step
reasoning and tool interactions. However, recent studies have revealed that LLM-based
agents can exhibit unfaithful behaviors, where their final reports fail to faithfully reflect their actual execution trajectories. Specifically, when
agents fail to accomplish tasks through the intended sources or procedures, they may resort to alternative sources or strategies while failing to disclose such deviations, leading users to incorrectly believe that the tasks have been completed according to the original requirements. We refer to this specific form of
agent unfaithfulness as
Agent Completion Unfaithfulness. Detecting and mitigating
agent completion unfaithfulness is critical yet remains largely unexplored. In this work, we propose MCAIR, a runtime monitoring and correction framework that leverages internal representations of LLM-based
agents. MCAIR uses multi-step internal representations as inputs to train a risk-aware probe, enabling online detection of potential completion unfaithfulness during execution. Furthermore, MCAIR introduces a risk-guided correction loop that
prompts agents to review their execution history and mitigate potential inconsistencies between their execution trajectories and final reports. Additionally, we construct ACUD, a
dataset for
Agent Completion Unfaithfulness detection that provides step-level execution traces, step-level internal representations, and consistency annotations for training and evaluating runtime detection probes. Experiments demonstrate that MCAIR outperforms LLM-as-a-judge baselines, generalizes across
prompts and tasks, and effectively reduces
agent completion unfaithfulness.
Read the full abstract ▾
All authors Yan Xu, Dandan Song, Luan Zhang, Jun Yang, Ke Duan, Yuhang Tian, Yading Li, Xudong Li, Xing Pei, Guangyuan Feng, Chenhao Li
Sound
TVPara: Paralinguistic Captioning of Expressive TV-Series Speech in Complex Acoustic Environments
Ziyang Zhang, Wenyi Yu, Wen Wu et al. (+2 more)
Large audio-
language models (LALMs) struggle to parse paralinguistics (e.g., emotion, pace, pitch) in
speech that combines intense emotion with complex acoustic environments. Strengthening this ability calls for training data that reflects such conditions. However, existing corpora rely heavily on studio-recorded
speech, offering limited emotional and acoustic diversity. To address this gap, we introduce TVPara, a paralinguistic captioning
dataset built from TV series.
Speech in TV series often carries diverse emotional expression and co-occurs with scene-specific background sound, providing expressive emotion and complex acoustic environments at scale. Constructed via LALM pre-annotation and human correction, TVPara includes two tasks: Single-Utterance (~492h) and Multi-Utterance (~213h) Paralinguistic Captioning. Fine-tuning open-source LALMs on TVPara yields an average relative improvement of 31.26% in paralinguistic understanding under complex acoustic environments. The gains further generalize to an unseen TV series and to cleaner, cross-domain
speech.
Read the full abstract ▾
All authors Ziyang Zhang, Wenyi Yu, Wen Wu, Xuenan Xu, Chao Zhang
Robotics
ACRO: Actor-Critic Rollout Orchestration
Junhoo Lee, Seungyeon Kim, Baekseung Kim et al. (+4 more)
Executing a
vision-
language-action (VLA) policy involves deciding when to continue an attempt and when to intervene to improve its chances of success. We introduce ACRO, an Actor–Critic rollout orchestration layer that makes these decisions for a fixed VLA. ACRO seeks interventions that offer a more promising continuation, are physically feasible, and leave sufficient time to complete the task. A Critic learned from policy rollouts estimates success prospects to decide when to intervene and select a re-entry target along the executed trajectory. Trajectory Retraction shortcuts the visited path to return the
robot to that target, where the same VLA resumes execution from a fresh observation. We evaluate ACRO with GR00T N1.7 and $\pi_{0.5}$ across SimplerEnv, RoboCasa, and real-
robot manipulation. Under the same physical execution budget, including all return motion, ACRO improves success by 10.4, 7.3, and 21.7 percentage points, respectively. It also completes additional tasks as the execution budget grows. These results show that controlling when and where a policy resumes can turn its existing capabilities into more successful task executions. Project page: https://junhoo.me/acro
Read the full abstract ▾
All authors Junhoo Lee, Seungyeon Kim, Baekseung Kim, Minkyu Kim, Suhyun Jeon, Jimyeong Kim, Nojun Kwak
Quantum Physics
Dynamical Decoupling of Nonclassical States in a Biased-Noise Spin Qudit
Arjen Vaartjes, Namit Anand, Rocky Y. Su et al. (+16 more)
Encoding
quantum information in $d$-dimensional systems (qudits) with $d>2$ offers appealing new directions for resource-efficient
quantum computing. However, protecting the coherence of nonclassical qudit resource states remains an open challenge. Dynamical decoupling (DD) is a well-established technique for extending coherence, but its application to qudits is hindered by the quadratic scaling of the decoupling sequence overhead with the dimension $d$. As a result, demonstrations of qudit DD have so far been limited to low-dimensional qudits. Here, we show that noise-biased qudits enable a nontrivial reduction of this overhead, for example, from quadratic to linear. We perform qudit DD on a model $d=8$ qudit, the nuclear
spin of an antimony donor in silicon. We
benchmark DD performance on highly nonclassical states such as uniform superpositions, cat states, and Haar random states and measure coherence improvements up to a factor $5$ to $10$. For a
spin-cat error correction code, continuous DD extends the coherence time of the logical superposition $|+\rangle_L$ to $225$~ms, compared to $19$~ms under free
evolution and $85$~ms with a Hahn echo. This improvement shows that DD sequences tailored to a system's noise bias can significantly enhance the lifetime of encoded logical
qubits.
Read the full abstract ▾
All authors Arjen Vaartjes, Namit Anand, Rocky Y. Su, Akram Youssry, Ritik Sareen, Jeffrey Marshall, Mark R. van Blankenstein, Danielle Holmes, Benjamin Wilhelm, Bibek Pokharel, Xi Yu, Laura A. O'Neill, Martin Nurizzo, Alexander M. Jakob, Fay E. Hudson, Kohei M. Itoh, Andrew S. Dzurak, David N. Jamieson, Andrea Morello
Quantum Physics
Transport Information Beyond Strong Eigenstate Thermalization
Avadhut V. Purohit, Harshit Sharma and Udaysinh T. Bhosale
Strong eigenstate thermalization (ETH) in operationally mean-ergodic, non-chaotic systems provides an opportunity to investigate whether Floquet eigenstates retain information about classical transport beyond their leading thermodynamic behavior. We address this question using the double kicked top (DKT), which exhibits nonmixing transport and dynamics consistent with weak mixing transport under variation of its control parameters. We show that the coefficient $m$ of the leading finite-size correction to the normalized eigenstate
entanglement entropy (EE) $s$ tracks the restructuring between these transport regimes even when the thermodynamic-limit
entropy $c$ becomes nearly insensitive to them. This coefficient $m$ is positive for nonmixing transport and negative when dynamics consistent with weak mixing transport dominate. The coefficient's sign reversal in the coexistence regime remains stable as the semiclassical resolution increases. Our results demonstrate that finite-size eigenstate structure can encode signatures of classical transport that are invisible in the thermodynamic-limit
entropy, providing an eigenstate-based diagnostic of dynamical organization beyond the leading thermodynamic behavior associated with strong ETH.
Read the full abstract ▾
Computer Vision
Remember, Revisit, Revise: Scaffold Memory for Long-Horizon 3D Geometry
Soumick Sarker and Mengmi Zhang
Geometry foundation models estimate camera poses and 3D scene structure from
video sequences, but maintaining consistent reconstructions over long horizons remains challenging. Feed-forward models process long sequences as independent, fixed-length frame windows and align their predictions, causing pose and reconstruction errors to accumulate across windows. Streaming models retain information from past frames to inform future predictions, but cannot revise earlier estimates when later observations reveal errors. To address this, we introduce Scaffold
MemOry for Revision (SMORE), a framework equipping frozen geometry foundation models with a revisable external
memory for long-horizon inference. Inspired by hippocampal associative
memory in biological
brains, which links spatial context with sensory experience, SMORE associates where an observer is with what it observes, enabling past observations to be stored, retrieved, verified against new evidence, and revised when inconsistencies are detected. We evaluate SMORE with eight competitive geometry models on five
datasets spanning camera pose estimation, multi-view depth estimation, 3D point-cloud reconstruction, simultaneous localization and mapping (SLAM), and novel view synthesis. Across all evaluated models, SMORE consistently improves long-sequence pose estimation, reduces multi-view depth error by up to 58%, reduces Chamfer distance for 3D point-cloud reconstruction by up to 43%, and achieves average trajectory error as low as 2.9 cm in long-horizon SLAM. By retaining corrected poses and observations, SMORE further improves the fidelity and spatial consistency of large-scale 3D scene reconstruction. These results establish revisable external
memory as a flexible interface for extending pretrained geometry models beyond fixed-context prediction toward iterative, error-correcting inference over long visual sequences.
Read the full abstract ▾
Probability
Exponential functionals of two random walks and stationary measures of discrete growth processes
Zikun Ouyang
We study random walks with
geometric or log-gamma increments, reweighted by certain exponential functionals. We establish precise asymptotics for all moments of exponential functionals, revealing a phase transition as the drifts of the walks vary. To describe the large time limit of reweighted random walks, we introduce a discrete $T$ transformation on random walk paths and establish Girsanov relations. As a corollary, we show that the stationary measures of
geometric last passage percolation and the log-gamma
polymer on a strip converge to their half-space counterparts as the strip width tends to infinity. In the log-gamma case, this provides an alternative proof of a recent result in arXiv:2607.17030.
Read the full abstract ▾
High Energy Astrophysical Phenomena
Constraint on particle mass of dark matter based on XTE J1814-338 observation and DANS model
Lin Wang and Dehua Wen
The small observed radius of
neutron star XTE J1814-338 (with mass of $1.21_{-0.05}^{+0.05}$ $M_{\odot}$ and radius of $7_{-0.4}^{+0.4}$ $\rm{km}$) poses a great challenge to the traditional
neutron star model composed of normal nuclear matter. The dark matter admixed
neutron star (DANS) model provides a clue to solve this challenge. By employing the self-interacting fermionic dark matter model, we investigate the effects of dark matter parameters including the particle mass, self-interaction strength, and mass fraction of dark matter on the properties of DANSs. Our results indicate that the joint constraints from the observation of XTE J1814-338 and the astrophysical limits on the dark matter self-interaction cross section derived from observations of different colliding
galaxy clusters can provide effective constraints on the particle mass of dark matter. Specifically, to simultaneously satisfy the small radius of XTE J1814-338 and the self-interaction limit, the particle mass of self-interacting fermionic dark matter is constrained to the range from sub-GeV to a few GeV.
Read the full abstract ▾
Solar and Stellar Astrophysics
Post-merger rotational evolution of magnetised single hot subdwarfs results in slow rotation due to magnetic braking
Facundo D. Moyano, Zhengwei Liu, Zhanwen Han et al. (+3 more)
Single hot subdwarfs are believed to form through mergers of helium-core white dwarfs, which can lead to large-scale
magnetic fields generated during the merger, and may be observable at the surface once they evolve into the stable core-helium burning phase. As a post-merger product, the orbital angular momentum of the binary system that forms it, could lead to a rapidly rotating hot subdwarf. But previous models were not able to explain the rather slow rotational velocities of observed single hot subdwarfs. Motivated by recent detections of surface
magnetic fields in hot subdwarfs in the range 200 - 500 kiloGauss, we explore in this work the role of large-scale
magnetic fields on their rotational
evolution via rotating stellar
evolution models. We find that large-scale stable
magnetic fields can extract angular momentum efficiently from the
star via coupling to stellar winds, producing slowly rotating single hot subdwarfs and thus providing a possible explanation to both the long rotation periods provided by asteroseismology and low spectroscopic rotational velocities. This hints
magnetic braking could be an important component of the rotational properties of low-mass stellar mergers.
Read the full abstract ▾
All authors Facundo D. Moyano, Zhengwei Liu, Zhanwen Han, Xuefei Chen, Tiara Battich, Philipp Podsiadlowski
Statistical Computation
SULA: Robust Langevin Sampling via Drift-Diffusion Saturation
Ritik Soni and Dootika Vats
The unadjusted Langevin
algorithm (ULA) and its
stochastic-gradient variant, SGLD, are workhorses of modern
Bayesian machine learning, but can be fragile when faced with a target distribution possessing a non-Lipschitz drift. In such settings, the expected squared jump distance can grow rapidly, leading to numerical instability or the need for prohibitively small step-sizes. We propose the saturated unadjusted Langevin
algorithm (SULA) that pairs a saturated gradient drift with a gradient-informed state-dependent
diffusion, while preserving first-order consistency with the Langevin
diffusion. The drift and
diffusion terms are chosen to guarantee that the one-step squared update is independent of the gradient. Consequently, SULA is robust to the choice of step-size and preserves estimation and prediction quality. We provide sufficient conditions under which SULA with a step-size $h$ converges weakly to the Langevin
diffusion at the standard $O(h)$ rate, the SULA chain is geometrically ergodic, and that the bias of its invariant measure relative to the target is $O(h)$. Further, we develop SG-SULA, a
stochastic-gradient version of SULA that rescales the
stochastic gradient using a Gaussian CDF correction under Gaussianity of the gradient noise. Numerical experiments on various examples demonstrate SULA's robustness to tuning and SG-SULA's improved predictive performance.
Read the full abstract ▾
Algebraic Geometry
Maximum Likelihood Degrees via One-Edge Chordal Completion
Nilava Metya
The maximum likelihood degree of a Gaussian graphical model measures the algebraic complexity of its likelihood equations. This paper studies maximum likelihood degrees of Gaussian graphical models through chordal completion. When a graph becomes chordal after one missing edge is added, its maximum likelihood degree is determined by the maximal cliques and clique-tree separators containing that edge. This recovers Uhler's formula for complete bipartite graphs with one part of size two, and extends it to deleted-spine clique books, whose degree is unchanged when a page is thickened to a clique. For several fill edges, the same method gives a product formula when no clique or separator bag contains two fill edges. We also determine four classes of vertices at which clique blow-up preserves the degree. Two follow from known results, and two are established here.
Read the full abstract ▾
High Energy Physics (Experiment)
Achieving a 0.49 keV threshold in the COSINE-100 dark matter experiment
J. Kim, N. Carlin, D. F. F. S. Cavalcante et al. (+58 more)
The COSINE-100 experiment searched for weakly interacting massive particles (WIMPs) with NaI(Tl)
crystals at the Yangyang Underground Laboratory. Using its full 6-year
dataset, we lower the analysis threshold to 0.49 keV. A multi-layer perceptron classifier separates scintillation events from photomultiplier tube noise. It is trained with approximately $3\times10^{6}$ simulated scintillation waveforms per
crystal, separately for groups of events with similar numbers of photoelectron clusters. The selection keeps more than 10% of scintillation events at 0.49 keV while limiting noise contamination to below 1%. The selection efficiency is validated with $^{22}$Na-calibration data and internal $^{22}$Na events, and the selection performance is stable over the six years of data taking. With this threshold, the median expected upper limit on the
spin-dependent WIMP-
proton cross section reaches $1.1\times10^{-36}$ cm$^2$ at a WIMP mass of 2.5 GeV/$c^2$, probing previously unexplored parameter space in the 2.0-2.8 GeV/$c^2$ mass range.
Read the full abstract ▾
All authors J. Kim, N. Carlin, D. F. F. S. Cavalcante, J. Y. Cho, S. J. Cho, S. Choi, A. C. Ezeribe, L. E. França, O. Gileva, C. Ha, I. S. Hahn, E. J. Jeon, H. W. Joo, W. G. Kang, M. Kauer, B. H. Kim, D. Y. Kim, H. J. Kim, K. W. Kim, S. H. Kim, S. K. Kim, W. K. Kim, Y. D. Kim, Y. H. Kim, B. R. Ko, Y. J. Ko, B. C. Koh, D. H. Lee, E. K. Lee, H. Lee, H. S. Lee, H. Y. Lee, I. S. Lee, J. Lee, J. Y. Lee, M. H. Lee, S. H. Lee, S. H. Lee, Y. J. Lee, D. S. Leonard, N. T. Luan, V. H. A. Machado, B. B. Manzato, R. H. Maruyama, R. F. Muhdi, S. L. Olsen, H. K. Park, H. S. Park, J. C. Park, J. S. Park, K. S. Park, K. Park, S. D. Park, R. L. C. Pitta, H. Prihtiadi, C. Rott, K. A. Shin, M. K. Son, L. T. Truc, L. Yang, G. H. Yu
Differential Geometry
Remarks on prescribed scalar curvature functions
Mingxiang Li, Quoc Anh Ngo and Xingwang Xu
This short article is devoted to the study of the positive solution of Yamabe-type semi-linear equation with prescribed coefficient on general closed
manifolds. Our main result is to provide some sufficient conditions for the existence of the positive solutions and give some examples to show that those conditions are sharp. An application to scalar-flat
manifold settles a long standing question in prescribing scalar
curvature problem in negative form.
Read the full abstract ▾
Group Theory
Invariable Generation of Groups of Homeomorphisms
Neelam, Davide Perego and Matteo Tarocchi
We provide a sufficient condition for a group of homeomorphisms of a compact Hausdorff space to be non-invariably generated, generalizing previous results proved using the dynamics of a group's action. We exhibit new applications among Scott-Röver-Nekrashevych groups and among groups of tree automorphism of with (almost) locally prescribed actions.
Read the full abstract ▾
Hardware Architecture
ESL-former: Event Sparse Linear Transformer for FPGA-Accelerated Malware Detection
Jun Yang, Yumeng Cao, Yingxuan Xiao et al. (+2 more)
Transformer based intrusion detection has shown strong potential for identifying unknown malware variants, but its deployment on resource constrained edge devices remains difficult due to the high computational and
memory cost of self
attention. This paper presents an event sparse Linear
Transformer framework for malware detection on FPGA edge platforms. Instead of relying on an explicit
attention matrix, the proposed method derives
token and channel importance directly from the accumulated key value state of Linear
Transformer. A contribution aware scoring function is introduced to measure the perturbation of each
token or kernel channel to the linear
attention state, allowing redundant
tokens and low contribution channels to be removed without constructing quadratic
attention maps. To translate sparsity into practical hardware acceleration, the proposed design further replaces conventional binary mask based execution with a Top K active
token buffer and a hardware compaction unit. The selected active
tokens and channels are reorganized into dense fixed length data streams, enabling downstream processing elements to execute regular multiply accumulate operations with reduced redundant
memory access and control overhead. This design is suitable for low power FPGA based malware detection, where both latency and on chip resource usage are critical. Experimental evaluation demonstrates that the proposed sparse Linear
Transformer can reduce hardware cost while preserving malware classification accuracy.
Read the full abstract ▾
All authors Jun Yang, Yumeng Cao, Yingxuan Xiao, Changhong Li, Shangru Zhou
Differential Geometry
The fence conjecture for convex projective polygons
Erxiao Wang
We prove the fence
conjecture of Dumas and Wolf: for a convex projective polygon with at least five vertices, every zero of the Pick differential lies in the interior of the region obtained by cutting off the triangles at consecutive triples of vertices. The proof gives a stronger estimate for arbitrary properly convex domains. In affine-sphere coordinates normalized at a Pick zero, the domain contains the closed disk of radius $2^{-1/4}$ and its closure lies in the open disk of radius $2^{1/4}$, unless it is a rotated square. The proof compares the canonical Monge-Ampère potential of the domain with that of a normalized square. Their third-order contact gives alternating local sign sectors. A planar separation argument and the maximum principle show that the negative sectors belong to four distinct components, each meeting a different side of the square. Rotation gives the inner disk bound, and conormal duality gives the outer bound. Both constants are optimal for the class of properly convex domains normalized at a Pick zero. The argument requires neither boundary smoothness nor any restriction on the multiplicity of the Pick zero.
Read the full abstract ▾
Artificial Intelligence
Learning When to Trust Medical Knowledge: Regression-Safe Routing for Clinical Reasoning
Hui Zhang, Yujia Zeng and Yuhao Xin
Clinical
reasoning LLMs need external medical knowledge, but retrieved evidence and
reasoning chains can be outdated, poisoned, or hallucinated. Harmful knowledge can cause regression: changing a correct answer into an incorrect one. We construct a medical knowledge-poisoning
benchmark covering retrieved evidence and
reasoning chains, with seven attack families filtered by a three-judge "subtle-but-wrong" vote and paired clean exams that separate poisoning effects from question difficulty. No fixed knowledge-use strategy is robust across these settings: strong RAG methods trade robustness to poisoned evidence against accuracy when evidence is useful. We formulate knowledge use as per-instance routing and introduce Med-TrustRouter, a lightweight controller that keeps the solver frozen. From leakage-free runtime signals, it estimates each action's expected correctness gain and uncertainty, preserving the base answer unless another action is confidently better. Across seven medical
reasoning settings, Med-TrustRouter achieves 78.9 macro and 69.2 worst-case accuracy, matching the strongest competitor on macro accuracy while substantially improving the worst-case floor. Its gains transfer across solvers and evidence judges, unlike a hand-coded router using the same signals. Similar performance across linear, tree, and neural action-value heads suggests that the benefit comes from routing rather than
estimator complexity. These results support learning when to trust external knowledge for robust clinical
reasoning.
Read the full abstract ▾
Numerical Analysis
C1 Virtual Element Methods for the Optimal Control of Oseen Equations with Stream-Function Formulation
Arbaz Khan, David Mora, Alberth Silgado et al. (+1 more)
This work analyzes a $C^1$-conforming virtual element method for the optimal control of generalized Oseen equations, formulated using the stream-function on simply connected polygonal domains. The method is based on a conforming discretization in $H^2$ of arbitrary order $k \ge 2$. The discrete scheme considers an approximation of the state and co-state equations which involves a specifically designed operator for the convective term and a variational discretization of the control variable. We derive quasi-optimal a priori error estimates in the $H^1$ and $H^2$-norms for the stream-function, and in the $L^2$-norm for the control, under standard assumptions on the family of polygonal meshes. The proposed method recovers the fluid velocity, vorticity, and pressure using suitable projections. Finally, we report several numerical experiments to support the theoretical findings using different polygonal meshes including the cavity problem. We highlight that this analysis also covers OCPs governed by Brinkman and Stokes equation.
Read the full abstract ▾
All authors Arbaz Khan, David Mora, Alberth Silgado, Harpal Singh
Fluid Dynamics
Response of Confined Mesoscale Flames to Acoustic Forcing
Akhil Aravind, Balasundaram Mohan and Saptarshi Basu
Mesoscale combustion plays a vital role in applications ranging from portable power generation to micro-propulsion systems. This study investigates the response of mesoscale flames to upstream acoustic forcing across three characteristic flame regimes: Flames with Repetitive Extinction and Ignition (FREI), Stable Flames (SF), and Propagating Flames (PF). Acoustic forcing is applied over a broad range of pressure intensities at the first harmonic of the combustor tube. At low forcing intensities, all the flame regimes exhibit oscillations at the forcing frequency without qualitative changes in their respective flame behaviours. However, as the acoustic pressure increases, all flame regimes transition to a common unsteady regime characterised by periodic ignition-extinction cycles (resembling FREI) alongside strong oscillations at the forcing frequency. This response is identified as the oscillating FREI (O-FREI) regime, in which increasing forcing intensity leads to a reduction in the mean heat release rate, an amplification of oscillations at the forcing frequency, and a drop in the Strouhal number associated with the ignition-extinction cycle. The response characteristics across all flame regimes are examined in detail and are further complemented using reduced-order mathematical models.
Read the full abstract ▾
Algebraic Geometry
Algebraically Integrable Smooth Convex Bodies
Junjie Shan
We prove that a convex body with $C^\infty$ boundary in an odd dimension $n\ge3$ is algebraically integrable if and only if it is an ellipsoid. This answers Arnold's problem for smooth convex bodies. Together with Vassiliev's even-dimensional nonexistence
theorem, it gives the classification in every dimension $n\ge2$. More generally, in odd dimensions, a convex body with $C^\infty$ semialgebraic boundary is an ellipsoid whenever its section volume function satisfies a joint algebraic relation.
Read the full abstract ▾
Computation and Language
Measuring and Reducing Unusable Replies of LLM Agents
Yilin Jiang, Fei Tan, Xuanyu Yin et al. (+3 more)
People delegate multi-step work to LLM
agents and read only the reply that the
agent hands back. Current evaluation scores task completion and the faithfulness of reports, and it leaves open whether the requester can understand and use the reply. We call this property reply usability and measure it with a reader model that takes the requester's position and tests whether the reply can be interpreted from what the requester has seen, is in the requester's
language and answers the request. Its agreement with the majority of five annotators lies within the range of the individual annotators. Across eight models from seven families, 54% of 480 replies fail, and the best model still fails on 30%. Failed replies mostly point to material the requester has not seen or use labels that the
agent coined during its work. Stating explicit requirements for the reply reduces the failure rate from 0.52 to 0.43 together with errors and omissions, and a secondary model that rewrites the reply reduces it to 0.36, which human annotators confirm in a blind comparison. Reply usability can thus be scored alongside task success and improved with one appended message or one call to a secondary model.
Read the full abstract ▾
All authors Yilin Jiang, Fei Tan, Xuanyu Yin, Guanchu Wang, Xia Hu, Yutao Yue
Complex Variables
Weighted composition–differentiation operators on mixed-norm, derivative Hardy and F(p,q,s) spaces
Jihua Sun and Junming Liu
We study weighted composition–differentiation operators on mixed-norm, derivative Hardy and F(p,q,s) spaces. Explicit dyadic criteria yield boundedness, compactness and two-sided norm and essential-norm estimates, including endpoint and quasi-Banach cases. We also obtain multiplier and approximation results, Schatten criteria on weighted Dirichlet spaces, and exact singular values for monomial weights and linear symbols.
Read the full abstract ▾
Computational Engineering
BioPRM: Step-Level Consistency Verification of Biological Protocols
Bowen Shao, Ziyang Ling, Ronald Xu et al. (+1 more)
Process reward models (PRMs) provide fine-grained supervision for assessing intermediate steps in
language-model outputs, but their application to biological protocols presents distinct challenges. Protocol validity depends on both the appropriateness of individual operations and their consistency with the experimental objective and surrounding procedure, while experimental outcomes are typically unavailable during verification. We introduce BioPRM, a contextual process reward model for identifying procedural inconsistencies in structured biological protocols. We construct a supervision corpus from SciRecipe and protocols collected from protocols.io and Bio-protocol using LLM-assisted construction and source-specific review by personnel with biological expertise.
Read the full abstract ▾
All authors Bowen Shao, Ziyang Ling, Ronald Xu, Mingzhai Sun
Robotics
Accelerating VLA Inference with Perturbation Informed Execution
Jehun Lee, Hajun Lee, Jimin Ryu et al. (+4 more)
Action-chunking
vision-
language-action (VLA) models predict sequences of future
robot actions. Under fixed-horizon execution, the
robot executes a fixedlength prefix of each predicted chunk before invoking the policy again, leaving the remaining predictions unused. For a given number of executed actions, short execution horizons require more policy calls. Extending the execution horizon amortizes policy inference cost over more actions. However, uniformly applying long horizons across episodes can limit opportunities for precise corrections. To address this limitation, we propose PICES (Perturbation-Informed Chunk and Execution Horizon Selection), a training-free method that jointly selects an action chunk and the number of actions executed before the next policy call. Within a single policy call, the action expert generates a batch of action chunks from multiple initial noise samples and their locally perturbed counterparts, using the same conditioning features across all samples. PICES exploits differences between corresponding actions in each nominal chunk and its counterpart to determine candidate-specific execution horizons and select a nominal chunk for execution. Experiments on simulation
benchmarks and real-world manipulation tasks demonstrate that PICES achieves speedup by up to 2.3x in wall-clock policy inference time on Jetson Orin and Thor relative to fixed horizon baselines while improving the success rate.
Read the full abstract ▾
All authors Jehun Lee, Hajun Lee, Jimin Ryu, Youhan Lee, Beomseok Kang, Jonghyun Choi, Jae-Joon Kim
Optimization and Control
Time-Inconsistent Singular Dividend Control with Regime Switching and Bankruptcy
Xiaochen Chen, Guohui Guan, Zongxia Liang et al. (+1 more)
This paper studies infinite-horizon
equilibrium dividend distribution for a firm whose surplus follows a
diffusion under observable Markov regime switching and is absorbed at bankruptcy. Dividend payments are modeled as singular controls, and non-exponential discounting leads to time inconsistency. We formulate
equilibrium control laws and establish a verification
theorem for a general finite-state model with surplus-dependent coefficients. For a two-state model with constant positive drifts and
volatilities and bi-exponential discounting, we prove the existence of a regime-dependent
equilibrium barrier policy for arbitrary positive switching intensities. The policy retains surplus below the prevailing regime's threshold and distributes any excess immediately. Numerical experiments examine how shareholder patience and economic conditions affect dividend decisions. Over the parameter ranges considered, greater discount-rate dispersion at a fixed initial average rate raises both barriers, while switching speed can generate nonmonotone responses and reverse their ordering.
Read the full abstract ▾
All authors Xiaochen Chen, Guohui Guan, Zongxia Liang, Xiaodong Luo
Computation and Language
HACHIMI-II: From Role-Play to Learner Behavior Modeling for Faithful Student Simulation
Yilin Jiang, Fei Tan, Xuanyu Yin et al. (+2 more)
Simulated students let researchers test tutoring systems and teaching strategies before real learners are available, but only if the simulation reproduces where a particular learner succeeds and fails. The common approach asks a large
language model to role-play a student, but a role does not model what this learner knows or will do next, so the model's own competence fills the gap. Simulated students perform at a level set by the model, differ little from one another, and seldom make the real learners' errors. We model the learner's behavior and recast student simulation as third-person prediction of the learner's next action. HACHIMI-II builds a student model from the learner's history, infers the learner's state for the current item or tutor turn, and compiles it into a behavior harness that constrains the next action. On three item-response
datasets, the Wasserstein-1 distance between simulated and real student ability is 2.4–8.4$\times$ smaller than that of the strongest baseline. On two dialogue
datasets, HACHIMI-II reproduces real student errors with the same knowledge type, locus, and mechanism 1.8–5.1$\times$ more often than the strongest baseline, and ablations trace the gains mainly to the explicit learner state. Predicting the learner makes capable models usable as simulated students.
Read the full abstract ▾
All authors Yilin Jiang, Fei Tan, Xuanyu Yin, Jing Leng, Aimin Zhou
Differential Geometry
Calabi metric on the space of k-Hessian potentials
Linwei Zhang
Let $(X,\omega)$ be a compact Kähler
manifold of complex dimension $n$, and let $1\leq k\leq n$. We consider the space of smooth $k$-admissible potentials introduced in the study of the Hessian-cscK equation by Guo–Smith–Tong. The complex $k$-Hessian equation identifies this space, modulo constants, with the space of smooth positive densities of fixed total mass. We compute the differential of the Hessian density map and use it to pull back the Fisher–Rao metric. The resulting Riemannian metric is \[ G^{\HC,k}_{\varphi}(u,v) =\int_X \bigl(\Delta_{\varphi}^{(k)}u\bigr) \bigl(\Delta_{\varphi}^{(k)}v\bigr) \,\omega_{\varphi}^{k}\wedge\omega^{n-k}. \] We then show that the square-root Hessian density map \[ \varphi\longmapsto 2\sqrt{\frac{\omega_{\varphi}^{k}\wedge\omega^{n-k}}{\omega^n}} \] is a global Riemannian isometry onto the positive part of the $L^2(X,\omega^n)$-sphere of radius $2\sqrt{V}$, where $V=\int_X\omega^n$. For $k=n$, this recovers the spherical description of the classical Calabi metric. We also derive the
curvature, geodesic and diameter formulas, and prove a quantitative local comparison between adjacent Hessian–Fisher distances.
Read the full abstract ▾
Machine Learning
On Global Joint Lower Bounds in Zeroth-Order Nonconvex Optimization
Yuxuan Xia and Xiao-Tong Yuan
Recent work has sought to transfer first-order nonconvex optimization guarantees to zeroth-order methods through gradient
estimators, typically incurring a factor linear in $d$. \citet{kornowski2024} obtained an $\mathcal{O}(d\delta^{-1}\epsilon^{-3})$ oracle complexity for nonsmooth nonconvex optimization and asked whether its dependence on all three parameters is jointly optimal. Although the dimension dependence and the first-order dependence on $(\delta,\epsilon)$ are separately known to be unavoidable, a matching zeroth-order lower bound has remained open. We answer this question in the unrestricted full-space setting. For smooth
stochastic nonconvex optimization, we establish a scalar-oracle lower bound of $\Omega\!\left(dL_0^2\min\{\Delta H,L_0^2\}/\epsilon^4\right)$ assuming a smooth, globally lower-bounded population objective and a second-moment bound on sample Lipschitz constants. A smooth reduction then yields a lower bound for Goldstein stationarity, proving the joint minimax optimality of the dependence on $(d,\delta,\epsilon)$ under explicit high-dimensional and small-accuracy conditions. Crucially, our results allow
algorithms to query arbitrary points in $\mathbb{R}^d$, adaptively select sample-function labels, reuse the same sampled function at any number of locations, and output points without spatial restrictions. Thus, unlike lower bounds based on bounded query regions or fresh evaluation noise, our guarantees apply to a fully adaptive, reusable-sample scalar oracle. We also establish a separate noiseless exact-value lower bound against randomized
algorithms, and hence also against deterministic methods. Our proof combines a globally defined sequential nonconvex construction with an exact Gaussian scalar channel, multistage random-field estimates, and an antithetic finite-sample realization that preserves information hardness under adaptive reuse.
Read the full abstract ▾